* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Structural elements defining elongation factor Tu mediated
Protein–protein interaction wikipedia , lookup
Gene expression wikipedia , lookup
Biochemical cascade wikipedia , lookup
Silencer (genetics) wikipedia , lookup
Metalloprotein wikipedia , lookup
Transcriptional regulation wikipedia , lookup
Paracrine signalling wikipedia , lookup
Proteolysis wikipedia , lookup
NADH:ubiquinone oxidoreductase (H+-translocating) wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Drug design wikipedia , lookup
Protein structure prediction wikipedia , lookup
Genetic code wikipedia , lookup
Biochemistry wikipedia , lookup
Amino acid synthesis wikipedia , lookup
Two-hybrid screening wikipedia , lookup
Ligand binding assay wikipedia , lookup
Artificial gene synthesis wikipedia , lookup
Epitranscriptome wikipedia , lookup
Biosynthesis wikipedia , lookup
3420–3430 Nucleic Acids Research, 2007, Vol. 35, No. 10 doi:10.1093/nar/gkm211 Published online 3 May 2007 Structural elements defining elongation factor Tu mediated suppression of codon ambiguity Hervé Roy, Hubert Dominique Becker, Marie-Hélène Mazauric and Daniel Kern* UPR 9002, ‘Architecture et Réactivité de l’ARN’, Institut de Biologie Moléculaire et Cellulaire du CNRS, 15, Rue René Descartes and Université Louis Pasteur, F-67084 Strasbourg Cédex, France Received January 23, 2007; Revised March 24, 2007; Accepted March 27, 2007 ABSTRACT Asn Gln In most prokaryotes Asn-tRNA and Gln-tRNA are formed by amidation of aspartate and glutamate mischarged onto tRNAAsn and tRNAGln, respectively. Coexistence in the organism of mischarged Asp-tRNAAsn and Glu-tRNAGln and the homologous Asn-tRNAAsn and Gln-tRNAGln does not, however, lead to erroneous incorporation of Asp and Glu into proteins, since EF-Tu discriminates the misacylated tRNAs from the correctly charged ones. This property contrasts with the canonical function of EF-Tu, which is to non-specifically bind the homologous aa-tRNAs, as well as heterologous species formed in vitro by aminoacylation of non-cognate tRNAs. In Thermus thermophilus that forms the Asp-tRNAAsn intermediate by the indirect pathway of tRNA asparaginylation, EF-Tu must discriminate the mischarged aminoacyl-tRNAs (aa-tRNA). We show that two base pairs in the tRNA T-arm and a single residue in the amino acid binding pocket of EF-Tu promote discrimination of Asp-tRNAAsn from Asn-tRNAAsn and Asp-tRNAAsp by the protein. Our analysis suggests that these structural elements might also contribute to rejection of other mischarged aa-tRNAs formed in vivo that are not involved in peptide elongation. Additionally, these structural features might be involved in maintaining a delicate balance of weak and strong binding affinities between EF-Tu and the amino acid and tRNA moieties of other elongator aa-tRNAs. INTRODUCTION In all living organisms, the translational machinery requires a minimum set of at least 20 aminoacyl-tRNAs (aa-tRNAs), one for each of the standard amino acids (aa) found in proteins. Most aa-tRNAs are formed by direct attachment of aa to the homologous tRNA by the cognate aminoacyl-tRNA synthetase (aaRS) (1,2). However, asparaginyl-tRNAAsn (Asn-tRNAAsn) and glutaminyltRNAGln (Gln-tRNAGln) can be formed via an alternate route in which tRNAAsn and tRNAGln are mischarged with aspartate (Asp) and glutamate (Glu), respectively, prior to amidation of the aa by a tRNA-dependent amidotransferase (AdT) (3–5). Mischarged Asp-tRNAAsn and Glu-tRNAGln are unable to bind in vitro elongation factor Tu (EF-Tu), which delivers aa-tRNAs to the ribosome A-site during peptide elongation (5,6). Nevertheless, poor binding of the mischarged AsptRNAAsn and Glu-ARNtGln was reported in vivo when the mischarged aaRS is overexpressed (7,8). It is proposed that because of the weak affinity of these mischarged aatRNAs for EF-Tu, the binding is low enough to prevent effective competition with the homologous aa-tRNAs under physiological conditions. Thus, these non-cognate aa-tRNAs escape protein synthesis, and therefore do not lead to misreading of codons. This property contrasts with the ability of EF-Tu to bind non-specifically to most aatRNAs including some non-cognate species formed in vitro (9,10) and tRNAs acylated with non-natural aa (11). Crystal structures of the EF-Tu GTP aa-tRNA ternary complexes (12,13) show that the protein recognizes aa-tRNA through non-specific contacts with the backbone of the tRNA and the aa moiety explaining, at the structural level, the wide specificity of EF-Tu for aatRNAs. To be delivered to the ribosome, aa-tRNAs must bind EF-Tu with an appropriate affinity (14). Recently, it has been shown that EF-Tu binds different cognate aatRNAs with a similar affinity that is maintained through a balance of strong and weak interactions with the aa and tRNA moieties of the aminoacylated tRNA (15). This suggests that the poor affinity of some misacylated tRNAs for EF-Tu results from weak binding interactions at both the aa and tRNA contact sites. However, the structural determinants that modulate EF-Tu binding affinity for the aa and tRNA moieties are not well understood. Asp-tRNAAsn and Glu-tRNAGln are two physiologically relevant mischarged aa-tRNAs that are *To whom correspondence should be addressed. Tel: þ33-3-8841-7092; Fax: þ33-3-8860-2218; Email: [email protected] Present address: Hervé Roy, Department of Microbiology, The Ohio State University, Columbus, OH 43210, USA ß 2007 The Author(s) This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/ by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Nucleic Acids Research, 2007, Vol. 35, No. 10 3421 unable to bind EF-Tu. This feature makes these aa-tRNAs well suited for investigations into structural determinants that modulate EF-Tu binding capacity. Here, we identify the tRNA and protein structural elements that determine the inability of Asp-tRNAAsn to bind EF-Tu. We show that two base pairs in the T-arm of the tRNA, which distinguish tRNAAsn and tRNAAsp, and a single residue in the aa binding pocket of EF-Tu are responsible for the loss of binding capacity. Surprisingly, these structural elements are conserved even in organisms that do not form Asp-tRNAAsn. Further, the tRNA elements that prevent binding of Asp-tRNAAsn to EF-Tu are present in other tRNA species that are excluded from protein synthesis as well as in some tRNAs involved in canonical aa-tRNA pairs, that do participate in peptide elongation. This suggests that the tRNA structural elements that are involved in the discrimination between Asp-tRNAAsn and Asp-tRNAAsp are likely also involved in the balance of binding interactions between tRNA and aa moieties of other cognate aa-tRNAs and EF-Tu (15). MATERIALS AND METHODS Strains, plasmids and proteins Thermus thermophilus EF-Tu and elongation factor Ts (EF-Ts), and Escherichia coli EF-Tu were overexpressed in E. coli (strains kindly provided by M. Sprinzl, University of Bayreuth and B. Kraal, University of Leiden) and purified as previously described (16,17). Mutagenesis of T. thermophilus EF-Tu was performed by PCR using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) with the pQE12 vector containing the EF-Tu ORF (18) and two complementary 35-mer oligonucleotides containing the desired mutations. Mutations were confirmed by DNA sequencing. The His-tagged EF-Tu variants were purified by precipitation of the thermolabile proteins from E. coli followed by chromatography on a Niþþ substituted Sepharose matrix, a procedure that prevents contamination by endogenous E. coli EF-Tu. tRNAAsn and tRNAAsp from T. thermophilus were purified from E. coli strain JM103 transformed with the pKK expression vector encoding the tRNA genes (19). The post-transcriptional modifications of the two tRNAs expressed in E. coli have not been characterized, but both are charged as efficiently and with the same rate by the homologous aaRS as the tRNAs isolated from T. thermophilus (not shown). Transcripts of wild-type and mutant T. thermophilus tRNAs and S. epidermidis tRNAGlyIA were obtained from reconstructed genes cloned into pUC18 as described (20). nitrocellulose discs and washed three times with 1 ml of buffer containing 10 mM Tris–HCl pH 7.4, 10 mM NH4Cl and 10 mM MgCl2. The% of active molecules of native and mutated EF-Tu, determined by nitrocellulose discs filtration of radioactive GDP bound on EF-Tu was evaluated to 70%. Multiple investigations have shown that molecules of T. thermophilus EF-Tu competent in the GDP exchange assay are active in formation of the ternary EF-Tu GTP aa-tRNA complex (16,18,21,22). Preparation of tRNA transcripts Prior to transcription, constructs were linearized by BstNI digestion to yield transcripts terminated by the 30 CCA. In vitro transcriptions were performed in 500 ml mixtures containing 40 mM Tris–HCl pH 8.1, 22 mM MgCl2, 1 mM spermidine, 5 mM DTE, 0.01% Triton X-100, 4 mM of each ATP, CTP, GTP and UTP, 16 mM GMP, 50 mg of linearized plasmid and 120 U of T7 RNA polymerase during 3 h of incubation at 378C. GMP was added to obtain transcripts starting with guanosine 50 monophosphate. To overcome the poor yield of transcription for tRNAAsn that starts with a U, the tRNAAsn gene was flanked at the 50 end by a cis-acting hammerhead ribozyme starting with three G residues to facilitate T7 polymerase recognition. Self-cleavage of the hammerhead ribozyme was completed during one hour of incubation at 608C after five-fold dilution of the transcription mix with 50 mM Tris–HCl pH 8.1 and 20 mM MgCl2 (23). Nucleic acids were recovered by ethanol precipitation, dried, dissolved in 1 ml water and applied to a Uno-Q6 (Bio-Rad, Hercules, CA, USA) anion-exchange chromatography column. Washing was performed with 40 ml of 50 mM Tris–HCl pH 7.5 followed by elution with a gradient from 0 to 1 M NaCl (2 25 ml). Transcripts eluted at 400 mM salt were precipitated with ethanol and dissolved in water. Their charging capacities were determined to be 60–70%. Preparation of aa-tRNAs Aminoacylation of tRNAs was performed in a mixture containing 4 mM transcript, 100 mM Na-HEPES pH 7.2, 30 mM KCl, 2 mM ATP, 10 mM MgCl2 and 20 mM of either L-[14C]Asp (270 cpm/pmol), L-[3H]Asn (500 cpm/pmol) or L-[14C]Phe (444 cpm/pmol) and 0.2 mM of T. thermophilus non-discriminating AspRS, AsnRS or PheRS. After 40 min incubation at 378C the reaction was stopped by addition of 60 mM sodium acetate pH 4.5 and 250 mM KCl and the aa-tRNAs were extracted with acid-buffered phenol and chloroform, and precipitated with ethanol. Pellets were dried, dissolved in water, and stored at 208C. Determination of the concentration of active EF-Tu Activation of EF-Tu The concentration of active EF-Tu was determined by active site titration as described previously (18,21). GDP complexed to EF-Tu was exchanged with [3H]-labeled GDP in a mix containing 50 mM Tris–HCl pH 7.4, 50 mM NH4Cl, 10 mM MgCl2, 2 mM [3H]GDP (500 cpm/pmol), 0.2 mM EF-Tu and 2 nM EF-Ts. After incubation for 20 min on ice, the reaction mix was filtered on Purified EF-Tu from T. thermophilus or E. coli (14 mM) were activated in a reaction mix containing 10 mM NaHEPES pH 7.2, 10 mM NH4Cl, 2 mM MgCl2, 1 mM GTP, 1 mM phosphoenolpyruvate, 5 U pyruvate kinase and 0.14 mM of T. thermophilus EF-Ts (22). Reactions were conducted for 20 min at 378C. Until use, the mix was stored on ice. 3422 Nucleic Acids Research, 2007, Vol. 35, No. 10 Filter binding assays for KD measurements Reaction mixtures of 400 ml contained 20 nM [14C]- or [3H]-labeled aa-tRNA and EF-Tu GTP in concentrations ranging from 0.02 to 2 mM in 50 mM Na-Hepes pH 7.2, 5 mM NH4Cl, 50 mM KCl, 10 mM MgCl2 and 1 mM GTP. After 5 min incubation at 48C, the mixture was filtered on a nitrocellulose disc presoaked in washing buffer. Filters were washed with 400 ml of reaction buffer depleted in GTP, dried and counted in a liquid scintillation counter. KD were determined by fitting the experimental data of concentrations of EF-Tu aa-tRNA complex, total EF-Tu and total aa-tRNA in the seconddegree equation derived from KD ¼ [free EF-Tu] [free aa-tRNA]/[EF-Tu aa-tRNA complex], in which [free EF-Tu] is substituted by [total EF-Tu] [EF-Tu aa-tRNA complex] and [free aa-tRNA] by [total aa-tRNA] [EF-Tu aa-tRNA complex]. The binding assays (Figure 1) show that the aa-tRNA bound to the EF-Tu GTP complex is totally retained on the nitrocellulose filter. Thus the efficiency of retention of the ternary complex is near 100%. The experimental data of [RL] (concentration of EF-Tu aa-tRNA), [Rt] (concentration of total EF-Tu) and [Lt] (concentration of total aa-tRNA) were fitted to the second degree equation whose positive solution expresses [RL] as a function of KD, [Rt] and [Lt]: [RL] ¼ pffi 1/2[(K D þ [Lt] þ [Rt]) ((K D þ [Lt] þ [Rt]) 2 4([Lt] [Rt]))] (24,25). KD and the correlation coefficient were calculated using the Kaleidagraph software. KD values are the average of at least three independent determinations where the KD of Asp-tRNAAsp was determined in parallel to serve as control. The affinity constants KA were calculated from 1/KD. The KD values we determined for Asp-tRNAAsp and Asn-tRNAAsn using the nitrocellulose filter-binding assay are one to two orders of magnitude higher than those determined by Ulenbeck’s group using the RNAse A protection assay (25). These variations may be related to the distinct methodologies used, or to subtle differences in the experimental conditions, since it has been shown that the KD values are strongly dependent upon the temperature and the ionic strength (15,25,26). To obtain comparable KD values each measurement involving tRNA and/or EF-Tu variants was conducted in parallel to determination of the KD of native EF-Tu GTP AsptRNAAsp and EF-Tu GTP Asn-tRNAAsn complexes. Therefore, absolute KD values obtained by our method might be less accurate than those obtained using the RNAse assay; however, using our comparative method, the losses or gains in binding efficiency for the various mutants that we measured are accurate. Kinetics of deacylation of aa-tRNA in the absence and in the presence of EF-Tu The reaction mixture of 100 ml contained: 25 mM Tris– HCl pH 7.4, 10 mM MgCl2, 35 mM KCl, 1 mM [14C] aa-tRNA and 5 or 15 mM activated T. thermophilus or E. coli EF-Tu (22). After 10 min of incubation at 48C, the mix was transferred to 378C and the aa-tRNA was determined after incubation times ranging from 0 to 60 min by trichloracetic acid precipitation of 15 ml aliquots. The half-lives of aa-tRNA (Ln 2/k) were determined from first order kinetics Ln (St/S0) ¼ kt, where St and S0 are the aa-tRNA at times t and t0 and k the rate constant of hydrolysis. Half-lives of 20 and 7 min were determined for free Asp-tRNAAsp and Asn-tRNAAsn, respectively. When the amount of aa-tRNA hydrolyzed in 60 min in the presence of EF-Tu could not be evaluated (less than 5%), its half-life was estimated to be increased more than 40-fold. RESULTS Base pairs G49-C65 and A51-U63 in the T-arm of tRNAAsn prevent binding of the aspartylated tRNA to EF-Tu Figure 1. Binding of aa-tRNAs to EF-Tu. Twenty nM T. thermophilus Asp-tRNAAsp, Asp-tRNAAsn and Asn-tRNAAsn were saturated with increasing amounts of T. thermophilus EF-Tu GTP. Bound complex retained on nitrocellulose filter discs was quantitated at described (see Materials and Methods section). Binding of aa-tRNA was specific since no radioactivity was retained in the absence of protein nor in the presence of bovine serum albumin. In the bacterium T. thermophilus, EF-Tu encounters two aspartylated tRNA species. The cognate species, AsptRNAAsp, is able to bind T. thermophilus EF-Tu while the mischarged Asp-tRNAAsn cannot (5). In this study, we investigated the structural elements of the protein and of the tRNAs involved in the discrimination by EF-Tu between the two aspartylated tRNAs. Nitrocellulose filter binding assays were used to determine the binding affinities of Asn-tRNAAsn, Asp-tRNAAsp, and of the mischarged Asp-tRNAAsn for EF-Tu (see Materials and Methods section). KD values of 270 and 110 nM were determined for Asn-tRNAAsn and Asp-tRNAAsp, respectively. However, the non-cognate Asp-tRNAAsn was deprived of significant affinity for EF-Tu (Figure 1). Since the same results were obtained with native purified tRNAAsp and tRNAAsn, the post-transcriptional modifications of these tRNAs do not contribute significantly to the discrimination of the aa-tRNAs by EF-Tu (data not shown). Nucleic Acids Research, 2007, Vol. 35, No. 10 3423 Figure 2. Strategy to identify nucleotides involved in Asp-tRNAAsn discrimination by EF-Tu. (A) Top left is the 3D structure of the EF-Tu GTP Phe-tRNAPhe complex (10) showing points of contact between tRNAPhe and EF-Tu. The secondary structure of Phe-tRNAPhe is shown at top right with tRNA nucleotides in the acceptor- and T-arms that contact EF-Tu represented by black spheres and numbered. A comparison of the acceptor and T-arms of tRNAAsp and tRNAAsn from T. thermophilus is shown at the bottom with nucleotides selected for mutational analysis boxed. (B) Comparison of base pairs 49–65 and 51–63 in tRNAAsp and tRNAAsn from various bacteria (44). The (þ) or () in front of the organism indicates whether the species uses (þ) the tRNA-dependent transamidation pathway for Asn-tRNAAsn formation or not (). To identify the tRNAAsn structural elements that determine the loss of binding of Asp-tRNAAsn to EF-Tu, we investigated the nucleotides that distinguish tRNAAsn (27) from tRNAAsp (28). Additionally, based on available crystal structure data, the nucleotides identified as candidates were restricted to those contacting the protein (12,13). According to these criteria, we identified five base pairs located within the acceptor- and T-stem regions of tRNAAsp and tRNAAsn (Figure 2A). We focused our analysis on base pairs 49–65 and 51–63 in the T-stem since: (i) structural elements of the T-stem were previously found to promote EF-Tu rejection of SectRNASec (29) and yeast Met-tRNAiMet (30); (ii) the crystal structure of the Cys-tRNACys EF-Tu GTP complex shows that nucleotide G63 establishes the unique specific interaction with residue E391 of the protein (13); and (iii) these two base pairs are well conserved in tRNAAsp and tRNAAsn throughout bacterial species (Figure 2B). Using site-directed mutagenesis we swapped the base pairs at positions 49–65 and 51–63, individually and in combination, between tRNAAsp and tRNAAsn. Purified transcripts of the tRNAAsp and tRNAAsn variants were aspartylated and the dissociation constants (KD) of the aa-tRNAs EF-Tu GTP complexes were determined using the filter-binding assays. Introduction in tRNAAsn of the G51-C63 base pair of tRNAAsp resulted in binding of mischarged Asp-tRNAAsn to EF-Tu, but with an affinity two orders of magnitude lower than Asp-tRNAAsp (Figure 3A). Similarly, introduction of the G49-U65 base pair of tRNAAsp into tRNAAsn also restored the EF-Tu binding capacity of Asp-tRNAAsn but with an affinity 50-fold lower than that of Asp-tRNAAsp (Figure 3A). Introduction of the two base pairs together in tRNAAsn promoted binding of Asp-tRNAAsn to EF-Tu with an affinity only 4-fold lower than that of Asp-tRNAAsp. Conversely, introduction in tRNAAsp of A51-U63 or G49-C65 from tRNAAsn led to a decreased affinity of Asp-tRNAAsp for EF-Tu by 3- and 6-fold, respectively, whereas introduction of both base pairs together decreased EF-Tu affinity by one order of magnitude. 3424 Nucleic Acids Research, 2007, Vol. 35, No. 10 Figure 3. Effects of mutation of base pairs 49–65 and 51–63 of tRNAAsp, tRNAAsn and tRNAPhe on affinity of aa-tRNAs for EF-Tu. (A) KD values of T. thermophilus EF-Tu for Asp- or Asn- tRNAAsn, Asp-tRNAAsp and variants. The acylating aa and nucleotides tested at positions 49–65 and 51–63 are indicated. Gray boxes represent nucleotides found in tRNAAsn and black boxes in tRNAAsp. (B) Ester-bond protection of wild-type AsptRNAAsp and Asp-tRNAAsn and variants in the presence of T. thermophilus EF-Tu GTP. Aa-tRNA protection is defined as the ratio of the half-life of aa-tRNA in the presence of EF-Tu GTP over that in its absence. (C) KD values of EF-Tu for wild-type Asp-tRNAPhe and variants. Base pairs in gray, black and white boxes represent those seen in wild-type tRNAAsn, tRNAAsp and tRNAPhe, respectively. (a),(b) Minimal value of KD determined with 340 nM of Asp-tRNAAsn and 10 mM of activated EF-Tu where 15%(a) or 30%(b) of aa-tRNAs were saturated. (c) Minimal value of KD determined with 20 nM of aa-tRNA and 2 mM of EF-Tu where 30% of aa-tRNA were saturated. (d) Less than 5% of the aa-tRNA ester bonds were hydrolyzed in 60 min (see Materials and Methods section). Surprisingly, the binding of Asn-tRNAAsn was mildly affected (3- to 5-fold) when the two base pairs (49–65 and 51–63) of tRNAAsp were introduced one at a time or in combination into tRNAAsn. This observation indicates that the effect of tRNA base pairs 49–65 and 51–63 on aa-tRNA binding by EF-Tu is strongly dependent on the aa acylating the tRNA. The importance of base pairs 49–65 and 51–63 of tRNA in binding of aa-tRNA on EF-Tu was also investigated by measuring the protection promoted by the protein of the aminoacyl ester bond of aa-tRNA against spontaneous hydrolysis (22,29). The half-life (t1/2) of aa-tRNA increases with the amount of aa-tRNA complexed with EF-Tu. Thus, protection of the aspartylated tRNAAsn variants was investigated by comparing the t1/2 in the presence and absence of EF-Tu (Figure 3B). EF-Tu binding to each aa-tRNA was estimated using the ratio of t1/2 in the presence of EF-Tu over that in its absence. We observed that while EF-Tu increases the half-life of Asp-tRNAAsp more than 30-fold, it has no effect on t1/2 of Asp-tRNAAsn. However, EF-Tu increased the half-life of the Asp-tRNAAsn variants containing the G49-U65 or the G51-C63 base pair by 2.4- and 7-fold, respectively, and that of the variant containing both base pairs by more than 30-fold. These results correlate with those obtained by direct binding measurements and establish the validity of the filter binding assays. Taken together, the earlier experiments demonstrate that for the tested aa-tRNAs, a G49-U65 base pair increases the affinity for EF-Tu while an A51-U63 base pair decreases it. Base pairs 51-63 and 49-65 of tRNA are sufficient to modulate the affinity of aa-tRNA for EF-Tu To determine whether base pairs 49–65 and 51–63 of tRNAAsp and tRNAAsn are sufficient to modulate the affinity of aspartylated tRNA for EF-Tu, we transplanted them into a different tRNA framework. Transplantation was performed into a T. thermophilus tRNAPhe variant that already contains aspartate identity elements for aminoacylation (20), and thus is able to be either phenylalanylated or aspartylated. Substitution of Nucleic Acids Research, 2007, Vol. 35, No. 10 3425 the base pairs 49–65 and 51–63 of tRNAPhe by the corresponding base pairs from tRNAAsp increased affinity of Asp-tRNAPhe for EF-Tu 10-fold, whereas substitution with those from tRNAAsn decreased its affinity by 12-fold (Figure 3C). Thus, the base pairs 51–63 and 49–65 modulate the affinity of aspartylated tRNAs for EF-Tu independently of the tRNA framework. Similar results were observed with phenylalanylated tRNA, but with a less pronounced effect. Incorporation of the base pairs from tRNAAsp increased affinity for EF-Tu by 1.4 fold, whereas introduction of those from tRNAAsn decreased the affinity by 2.5-fold. These results indicate that the magnitude of the observed effect, in response to 51–63 and 49–65 base pairs substitutions, is dependent on the nature of the acylating aa. The observed effects from nucleotide substitutions in tRNA acylated with Phe were weaker than those for tRNA acylated with Asp. This agrees with previous studies showing that the binding of aa-tRNAs by EF-Tu involves a synergistic recognition of both the aa and tRNA moieties (15). E. coli EF-Tu discriminates Asp-tRNAAsn from Asp-tRNAAsp Similar to what was observed with the protein from T. thermophilus, EF-Tu from E. coli did not bind Asp-tRNAAsn nor protect the aminoacyl ester bond of this aa-tRNA from spontaneous hydrolysis (Figure 4). This result is of particular interest, because it demonstrates the conservation of the discrimination properties of EF-Tu in E. coli, even though this organism does not use the indirect pathway of Asn-tRNAAsn formation and therefore is not expected to form Asp-tRNAAsn (5). Moreover, analysis of tRNAAsp and tRNAAsn sequences from various bacterial species reveals that the 49–65 and 51–63 base pairs are conserved with few exceptions, regardless of whether the tRNA-dependent transamidation pathway is used or not (Figure 2B). Consequently, the ability of EF-Tu to discriminate Asp-tRNAAsn from Asp-tRNAAsp and Asn-tRNAAsn is conserved even in organisms that do not require this property. bacterial elongation factors and are located 3.7 and 5.4 Å, respectively, from the aa side-chain in the crystal structure of the Phe-tRNAPhe EF-Tu GTP ternary complex (Figure 5B) (12). To analyze their role in discrimination of Asp-tRNAAsn by EF-Tu, both residues were replaced independently by Ala. The EF-Tu D228A mutant binds both Asp-tRNAAsp and Asn-tRNAAsn with affinities similar to wild type EF-Tu, but is still unable to bind Asp-tRNAAsn (Figure 5C). However, the E227A mutant binds mischarged Asp-tRNAAsn with an affinity only 5- and 2-fold lower than that of the wild-type EF-Tu for Asp-tRNAAsp and Asn-tRNAAsn, respectively. This indicates that the conserved E227 in the aa-binding pocket of EF-Tu is sufficient to promote rejection of Asp-tRNAAsn. Interestingly, the E227A mutation also slightly increases the affinity of EF-Tu for Asp-tRNAAsp and Asn-tRNAAsn, but does not significantly affect the affinity for Phe-tRNAPhe, indicating that the observed effect depends on the nature of the tRNA and the aa moieties. To determine if E227 excludes binding of Asp-tRNAAsn through electrostatic repulsion of the negatively charged Asp, we analyzed the binding properties of an E227Q mutant of EF-Tu. This variant, like wild-type EF-Tu, binds Asp-tRNAAsp and Asn-tRNAAsn, but is unable to bind Asp-tRNAAsn (Figure 5C). Since, either Glu or Gln at position 227 prevents binding of Asp-tRNAAsn to EF-Tu, rejection of this aa-tRNA is probably due to steric hindrance between the two aa side chains rather than to electrostatic repulsion. All archaea use the indirect pathway to form Gln-tRNAGln and about half of them to form Asn-tRNAAsn (31,32). Alignment of EF-Tu and archaeal EF-1A sequences reveals the presence, in the archaeal EF-Tu residue E227 is responsible for discrimination of naturally mischarged Asp-tRNAAsn The observation that aa-tRNAAsn binds EF-Tu when the esterifying aa is Asn and not Asp, suggests that the aa binding-pocket of the protein displays structural elements that selectively discriminates Asp when bound to tRNAAsn. Asp contains a negative charge within its sidechain that is not present in Asn. One possibility is that an acidic residue within EF-Tu in close proximity to the aa acylating the tRNA, may exert electrostatic repulsion on the negatively charged side-chain of Asp resulting in a weak interaction at the aa moiety. In the context of a weakly binding tRNA moiety, this repulsion could result in loss of binding to EF-Tu as seen in Asp-tRNAAsn. Sequence alignments of EF-Tu aa-binding pockets revealed two negatively charged residues (Figure 5A). These residues, Glu and Asp (E227 and D228 in T. thermophilus EF-Tu) are strictly conserved in all Figure 4. Protection of the ester bonds of Asp-tRNAAsp and AsptRNAAsn by EF-Tu. The kinetics of hydrolysis of the ester bond of T. thermophilus Asp-tRNAAsp (circles) and Asp-tRNAAsn (squares) were determined in the presence of E. coli EF-Tu GTP (filled symbols) or in its absence (open symbols) as described in Materials and Methods section. The curves Ln (St/S0) ¼ kt are shown where St and S0 are the concentrations of aa-tRNA at times t and t0, and k is the rate constant of hydrolysis of aa-tRNA. 3426 Nucleic Acids Research, 2007, Vol. 35, No. 10 factor, of a Gln residue at position corresponding to E227 in T. thermophilus EF-Tu (Figure 5A). This suggests that, in archaea using the tRNA-dependent transamidation pathway for Asn-tRNAAsn synthesis (31,32), this residue may prevent binding of the elongation factor to the Asp-tRNAAsn intermediate. Additionally, a Gln at this position is also conserved in eukaryal EF-1A. Interestingly, the tRNA-dependent transamidation pathway, and thus the necessity to discriminate between Asp-tRNAAsn and Asp-tRNAAsp, has so far not been shown to be necessary in eukaryotes. EF-Tu residue E391 modulates the affinity of EF-Tu for Asp-tRNAAsp and Asn-tRNAAsn Crystal structures of Cys-tRNACys or Phe-tRNAPhe in complex with T. thermophilus EF-Tu GTP show that recognition of aa-tRNAs by EF-Tu is mediated through non-specific hydrogen bonds at the backbone of tRNAs. However, one specific interaction in the Cys-tRNACys EF-Tu GTP complex was reported between the Og of EF-Tu residue E391 and the N2 of base G63 located in the minor groove of the C51-G63 base pair in the T-arm of the tRNA (13). The Glu at position 391 is strictly conserved in eubacterial EF-Tu. To investigate the importance of this residue for EF-Tu binding to Asp-tRNAAsp and Asn-tRNAAsn, we mutated it into Ala. The E391A EF-Tu variant displayed an increased affinity (2.8-fold) for Asn-tRNAAsn and a decreased affinity (5.5-fold) for Asp-tRNAAsp (data not shown). These results indicate that residue E391 modulates the affinity of EF-Tu for these aa-tRNAs. The G51-C63 base pair of tRNAAsp provides an NH2 group in the minor groove (33) like in the Cys-tRNACys EF-Tu GTP crystal structure, that may also contact residue E391 of EF-Tu by a specific hydrogen bond. Replacement of the Asp-tRNAAsp base pair G51-C63 with A51-U63, removes the hydrogen bond donor group, which might explain how this substitution decreases the affinity of Asp-tRNAAsp for EF-Tu. In contrast, AsntRNAAsn already contains an A51-U63 base pair, which does not provide a hydrogen bond donor group for interaction with EF-Tu residue E391. This might explain why the E391A mutation of EF-Tu does not negatively affect binding of Asn-tRNAAsn. In fact, a slight increase in affinity is observed between the E391A variant and this aatRNA. This structural interpretation to explain the variable affinity of the tRNA moiety of aa-tRNA for the elongation factor is also supported by the base pair swapping experiments between tRNAAsp and tRNAAsn described earlier. Replacement of the base pair G51-C63 by A51-U63 in tRNAAsp decreases the affinity of the aspartylated tRNAAsp for EF-Tu. Conversely, replacement of A51-U63 by G51-U63 in tRNAAsn increases the affinity of aspartylated or asparaginylated tRNAAsn for EF-Tu (Figure 3). DISCUSSION Base pairs 49–65 and 51–63 of tRNA modulate the affinity of aa-tRNAs for EF-Tu Figure 5. Identification of aa residues in the aa-binding pocket of EF-Tu potentially involved in rejection of Asp-tRNAAsn. (A) Sequence alignment of the region encompassing E227 and D228 of bacterial EF-Tu and comparison with archaeal and eukaryal EF-1A. A sampling of 87 bacterial elongation factors is shown as well as the consensus sequences of bacterial EF-Tu, and archaeal and eukaryal EF-1A. The percentage of homology is symbolized by colors: red and orange denote 100 and 80% identity, respectively, and yellow indicates conservation of the chemical nature of the residue in at least 80% of the sequences. (B) Localization of E227 and D228 in the aa-binding pocket of T. thermophilus EF-Tu. The structure displayed is that of the EF-Tu GTP Phe-tRNAPhe ternary complex (12). The two acidic residues that belong to the aa-binding pocket are in red and the tRNA is in purple. (C) Dissociation constants of wild-type T. thermophilus EF-Tu and mutants with specified aa-tRNAs. a Minimal value of KD estimated with 340 nM of Asp-tRNAAsn and 10 mM of activated EF-Tu where 15% of aa-tRNAs were saturated. We have shown that the distinct binding strengths of tRNAAsp and tRNAAsn for EF-Tu are promoted by base pairs 49–65 and 51–63 located in the T-arm. One question emerging from this study is whether these elements constitute the basis for EF-Tu discrimination of other aa-tRNAs escaping peptide elongation. Like Asp-tRNAAsn, the Glu-tRNAGln intermediate, formed by the indirect Gln-tRNAGln transamidation pathway (4,34,35), does not bind EF-Tu (36). Comparison of consensus sequences from bacterial tRNAGln and tRNAGlu shows that tRNAGln does not contain a U in the 49–65 base pair, but contains a U in the 51–63 pair (Figure 6A). In contrast, tRNAGlu contains a U in the first pair but not in the second one. With only a few exceptions, these base pairs are conserved in tRNAGln and tRNAGlu, and they are generally present regardless of whether or not the organism uses the indirect pathway of Gln-tRNAGln formation. Thus, EF-Tu may discriminate the mischarged and Asp-tRNAAsn using similar Glu-tRNAGln mechanisms. Our results, in combination with a previous work (26), suggest that Asp and Glu have a weak affinity for EF-Tu and that for efficient binding of aspartylated or glutamylated tRNA to the elongation factor, the tRNA moiety Nucleic Acids Research, 2007, Vol. 35, No. 10 3427 should display a strong binding ability provided by the G49-U65 and G51-C63 base pairs. It is worth noting that a G-U pair provides unique structural features. This wobble base pair induces a deviation of up to 2 Å five base pairs away (37) and provides a peculiar local electrostatic environment (38). These structural properties may influence aa-tRNA binding to EF-Tu when a wobble base pair is embedded in the T-stem of the tRNA. Base pairs G49-U65 and G51-C63 are conserved in elongator tRNAGly and absent in non-elongator tRNAGly Staphylococcus epidermidis contains in addition to the elongator tRNAGly, the non-elongator species, Gly-tRNAGlyIA, which is not involved in protein synthesis, but instead serves for synthesis of cell wall peptidoglycan (39,40). Like tRNAAsp, elongator tRNAGly contains the G49-U65 and G51-C63 pairs that may confer a strong binding capacity of Gly-tRNAGly for EF-Tu, compensating the low affinity of the Gly moiety. In contrast, the affinity of the glycylated native and transcript tRNAGlyIA for EF-Tu is at least one order of magnitude lower (data not published) (41). Interestingly, tRNAGlyIA contains A–U pairs at positions 49–65 and 51–63, which might explain the absence of recognition by EF-Tu (Figure 6A). This idea is strongly supported by a recent study showing that replacement of the base pair G49-U65 by G49-C65 drastically decreases the elongator capacity of E. coli tRNAGly in vivo (42). The formylated initiator Met-tRNAMet (fMet-tRNAiMet), which binds to the initiator factor IF2 constitutes another example of aatRNA deprived of affinity for EF-Tu. It has been shown that lack of binding of fMet-tRNAiMet to E. coli EF-Tu is promoted by the formyl group of Met and by the base pair C1-A72, which acts as a negative determinant, since substitution of this pair by C1-G72 or U1-A72 improves binding (43). These examples and the results reported here illustrate that there might be alternative mechanisms and multiple structural determinants responsible for either strong or weak binding interactions between aa-tRNAs and EF-Tu. The frequency of U in base pairs 49–65 and 51–63 correlates with the affinities of other tRNAs for EF-Tu To be delivered to the ribosome, elongator aa-tRNAs must bind EF-Tu with appropriate and equivalent affinities, which are conferred by a thermodynamic compensation between a strong interaction of the aa moiety and a weak interaction of the tRNA moiety or vice-versa. The contributions of the aa and tRNA moieties to the overall affinity of aa-tRNA for EF-Tu are organized in a compensatory fashion such that all elongator aa-tRNAs bind with similar affinities (14,15,26,44). In a previous study, the contribution of the tRNA moieties in binding of aa-tRNAs to EF-Tu were evaluated by affinity measurements of various E. coli tRNAs charged with Val (26,44). To analyze whether the tRNA base pairs at positions 49–65 and 51–63 influence the binding strength of these tRNAs, we investigated the correlation between the nature of the residues at these positions in tRNAs and the affinity of the tRNAs for EF-Tu. Figure 6 shows that E. coli tRNAs fall into two categories according to their relative affinities for EF-Tu. tRNAs exhibiting the strongest affinity (G0 ¼ 11.7 to 10.5 kcal/mol) all possess a U in the 49–65 base pair (except tRNAThr) and are deprived of a U in the 51–63 pair. The majority of tRNAs exhibiting lower affinity for EF-Tu (G0 ¼ 9.6 to 8.5 kcal/mol) do not contain a U within the 49–65 base pair but contain a U in the 51–63 base pair. To analyze whether these nucleotides are conserved in homologous tRNAs from other species (45), we compared the frequency of nucleotides which determine strong affinity of tRNA for EF-Tu to that of nucleotides which determine low affinity in other organisms. tRNAs homologous to those of E. coli that exhibit strong EF-Tu binding capacity combine a high frequency of U65 and a notable absence of U in position 63. Conversely, tRNAs homologous to those of E. coli that exhibit low EF-Tu binding capacity combine a high frequency of U in position 63 and an absence of U at position 65 (Figure 6). These observations suggest that the structural elements of tRNA that determine the strong affinity of tRNAAsp and the weak affinity of tRNAAsn for EF-Tu are, at least partly, conserved in other tRNAs in E. coli as well as in other organisms. Discrimination of Asp-tRNAAsn allows unambiguous decoding of Asn codons Current evolutionary models agree that AsnRS and GlnRS, which promote the direct pathway for aminoacylation of tRNAAsn and tRNAGln, are a recent addition to the repertoire of aaRSs (46–48). Introduction of Asn and Gln in the genetic code is believed to have occurred earlier in the course of evolution through the use of the indirect pathway of synthesis of Asn-tRNAAsn and Gln-tRNAGln. This ancestral pathway relies on the tRNA-dependent amidotransferase, the enzyme that converts misacylated Asp-tRNAAsn and Glu-tRNAGln into Asn-tRNAAsn and Gln-tRNAGln. The tRNA-dependent biosynthesis pathway of these aa is viable only if the mischarged intermediates (Asp-tRNAAsn and Glu-tRNAGln) are efficiently excluded from protein synthesis, in order to avoid ambiguous decoding of Asn and Gln codons. This suggests that the ability of EF-Tu to discriminate misacylated Asp-tRNAAsn and Glu-tRNAGln preexisted or appeared concomitantly to the introduction of Asn and Gln into the genetic code. The structural elements involved in the discrimination of Asp-tRNAAsn are widely conserved in eubacterial tRNAAsn and EF-Tu, even in organisms that do not use the indirect pathway of tRNA asparaginylation. In these organisms, the formation of mischarged Asp-tRNAAsn is not expected to occur, even though the structural elements that allow EF-Tu to discriminate the mischarged aatRNA are still present. The analysis of the base pair composition of elongator tRNAs (see earlier and Figure 6), suggests that these elements may play a larger role than just discrimination of non-cognate aa-tRNAs and may be involved in the modulation of EF-Tu binding to other elongator aa-tRNAs. The maintenance of this delicate balance of interactions is essential for the 3428 Nucleic Acids Research, 2007, Vol. 35, No. 10 Figure 6. Analysis of base pairs 49–65 and 51–63 in the T-stem of various eubacterial tRNAs. (A) Comparison of base pairs 49–65 and 51–63 of aatRNAs that are discriminated or recognized by EF-Tu. The reported base pairs are conserved in more than 80% of specified tRNAs. tRNAGlyIA base pairs are those from S. aureus and S. epidermidis tRNAs (B) Correlation between the Go of formation of EF-Tu GTP Val-tRNA complexes and the frequency of U in base pairs 49–65 and 51–63. The frequency of U in the 49–65 pair in combination with absence of U in the 51– 63 pair is represented by black bars. The frequency of U in the 51–63 pair and absence of U in the 49–65 pair is represented by dark gray bars. The percentage of the conservation of the combinations described above is on the left axis. The Go of formation of the complex of valylated tRNAs with EF-Tu (26,44) is represented by white bars and is indicated on the right axis. The specificities of the tRNAs are indicated by the single letter aa code. The sequences of base-pairs 49–65/50–64/51–63 of E. coli tRNAs and the number of eubacterial tRNA species analyzed are indicated at the bottom of the Figure. tRNA sequences were obtained from the compilation of tRNA sequences and genes (49). role of EF-Tu to discriminate between cognate and noncognate aa-tRNAs and properly deliver the correct substrate to the translational machinery for incorporation during protein synthesis. la Recherche et de la Technologie. Funding to pay the Open Access publication charges for this article was provided by the CNRS. Conflict of interest statement. None declared. ACKNOWLEDGEMENTS The authors thank Dr B. Kraal (University of Leiden, The Netherlands), Pr M. Sprinzl (University of Bayreuth, Germany) for the E. coli strains overexpressing E. coli and T. thermophilus EF-Tu and A.M. Smith and Dr M. Ibba (The Ohio State University, Columbus, OH, USA) for critical reading of the manuscript. This work was supported by the Université Louis Pasteur (Strasbourg) and by the Centre National de la Recherche Scientifique (CNRS) and by grants from the Association pour la Recherche sur le Cancer (ARC) and ACI Biologie Moléculaire et Structurale. H.R. was a recipient of a fellowship from the Ministère de l’Education Nationale de REFERENCES 1. Söll,D. and Schimmel,P.R. (1974) Aminoacyl-tRNA synthetases. Enzyme, 10, 489–538. 2. Ibba,M., Becker,H.D., Stathopoulos,C., Tumbula,D.L. and Söll,D. (2000) The adaptor hypothesis revisited. Trends Biochem. Sci., 25, 311–316. 3. Wilcox,M. (1969) Gamma-glutamyl phosphate attached to glutamine-specific tRNA. A precursor of glutaminyl-tRNA in Bacillus subtilis. Eur. J. Biochem., 11, 405–412. 4. Curnow,A.W., Hong,K., Yuan,R., Kim,S., Martins,O., Winkler,W., Henkin,T.M. and Söll,D. (1997) Glu-tRNA(Gln) amidotransferase: a novel heterotrimeric enzyme required for correct decoding of glutamine codons during translation. Proc. Natl Acad. Sci. USA, 94, 11819–11826. Nucleic Acids Research, 2007, Vol. 35, No. 10 3429 5. Becker,H.D. and Kern,D. (1998) Thermus thermophilus: A link in the evolution of the tRNA-dependent amino acid amidation pathways. Proc. Natl Acad. Sci. USA, 95, 12832–12837. 6. Curnow,A.W., Tumbula,D.L., Pelaschier,J.T., Min,B. and Söll,D. (1998) Glutamyl-tRNA(Gln) amidotransferase in Deinococcus radiodurans may be confined to asparagine biosynthesis. Proc. Natl Acad. Sci. USA, 95, 12838–12843. 7. Baik,J.-W., Yoon,J.-H., Namggong,S., Söll,D. and Kim,S.-I. (2004) Growth inhibition of Escherichia coli during expression of Bacillus subtilis glutamyl-tRNA synthetase catalyzes the formation of mischarged glutamyl-tRNAGln. J. Microbiol., 42, 111–116. 8. Min,B., Kitabatake,M., Polycarpo,C., Pelaschier,J., Raczniak,G., Ruan,B., Kobayashi,H., Namggong,S. and Söll,D. (2003) Protein synthesis in Escherichia coli with mischarged tRNA. J. Bact., 185, 3524–3526. 9. Chapeville,F., Lipman,F., Ehrenstein,G.V., Weisblum,B., Ray,W.H. and Benzer,S. (1962) On the role of soluble ribonucleic acid in coding for amino acids. Proc. Natl Acad. Sci. USA, 48, 1086–1092. 10. Mulvey,R.S. and Fersht,A.R. (1977) Editing mechanisms in aminoacylation of tRNA: ATP consumption and the binding of aminoacyl-tRNA by elongation factor Tu. Biochemistry, 16, 4731–4737. 11. Wang,L., Zhang,Z., Brock,A. and Schultz,P.G. (2003) Addition of the keto functional group to the genetic code of Escherichia coli. Proc. Natl Acad. Sci. USA, 100, 56–61. 12. Nissen,P., Kjeldgaard,M., Thirup,S., Polekhina,G., Reshetnikova,L., Clark,B.F. and Nyborg,J. (1995) Crystal structure of the ternary complex of Phe-tRNA(Phe), EF-Tu, and a GTP analog [see comments]. Science, 270, 1464–1472. 13. Nissen,P., Thirup,S., Kjeldgaard,M. and Nyborg,J. (1999) The crystal structure of Cys-tRNACys-EF-Tu-GDPNP reveals general and specific features in the ternary complex and in tRNA. Structure, 7, 143–156. 14. Ibba,M. (2001) Protein synthesis. Discriminating right from wrong. Science, 294, 70–71. 15. LaRiviere,F.J., Wolfson,A.D. and Uhlenbeck,O.C. (2001) Uniform binding of aminoacyl-tRNAs to elongation factor Tu by thermodynamic compensation. Science, 294, 165–168. 16. Blank,J., Grillenbeck,N.W., Kreutzer,R. and Sprinzl,M. (1995) Overexpression and purification of Thermus thermophilus elongation factors G, Tu, and Ts from Escherichia coli. Protein Expr. Purif., 6, 637–645. 17. Boon,K., Vijgebboom,E., Madsen,L.V., Talens,A., Kraal,B. and Bosch,L. (1992) Isolation and functionnal analysis of His-tagged elongation factor Tu. FEBS Lett., 210, 177–183. 18. Ribeiro,S., Nock,S. and Sprinzl,M. (1995) Purification of aminoacyl-tRNA by affinity chromatography on immobilized Thermus thermophilus EF-Tu.GTP. Anal. Biochem., 228, 330–335. 19. Roy,H. (2002) Evolution des voies de formation de l0 asparagine et d’asparaginylation de l’ARNt. Thesis, University Louis Pasteur, Strasbourg France. 20. Becker,H.D., Giegé,R. and Kern,D. (1996) Identity of prokaryotic and eukaryotic tRNA(Asp) for aminoacylation by aspartyltRNA synthetase from Thermus thermophilus. Biochemistry, 35, 7447–7458. 21. Miller,D.L. and Weissbach,H. (1974) Elongation factor Tu and the aminoacyl-tRNA-EFTu-GTP complex. Methods Enzymol., 30, 219–232. 22. Rudinger,J., Blechschmidt,B., Ribeiri,S. and Sprinzl,M. (1994) Minimalist aminoacylated RNAs as efficient substrates for elongation factor Tu. Biochemistry, 33, 5682–5688. 23. Fechter,P., Rudinger,J., Giegé,R. and Théobald-Dietrich,A. (1998) Ribozyme processed tRNA transcripts with unfriendly internal promoter for T7 RNA polymerase: production and activity. FEBS Lett., 436, 99–103. 24. Pleiss, J.A. and Uhlenbeck, O.C. (2001) Identification of thermodynamically relevant interactions between EF-Tu and backbone elements of tRNA. J. Mol. Biol. 308, 895–905. 25. Asahara,H. and Uhlenbeck,O.C. (2005) Predicting the binding affinities of misacylated tRNAs for Thermus thermophilus EF-Tu GTP. Biochemistry, 44, 11254–11261. 26. Asahara,H. and Uhlenbeck,O.C. (2002) The tRNA specificity of Thermus thermophilus EF-Tu. Proc. Natl Acad. Sci. USA, 99, 3499–3504. 27. Becker,H.D., Roy,H., Moulinier,L., Mazauric,M.H., Keith,G. and Kern,D. (2000) Thermus thermophilus contains an eubacterial and an archaebacterial aspartyl-tRNA synthetase. Biochemistry, 39, 3216–3230. 28. Keith,G., Yusupov,M., Briand,C., Moras,D., Kern,D. and Brion,C. (1993) Sequence of tRNA(Asp) from Thermus thermophilus HB8 [published erratum appears in Nucleic Acids Res 1993 Sep 25;21(19):4658]. Nucleic Acids Res., 21, 4399. 29. Rudinger,J., Hillenbrandt,R., Sprinzl,M. and Giegé,R. (1996) Antideterminants present in minihelix(Sec) hinder its recognition by prokaryotic elongation factor Tu. Embo J., 15, 650–657. 30. Kiesewetter,S., Ott,G. and Sprinzl,M. (1990) The role of modified purine 64 in initiator/elongator discrimination of tRNAMeti from yeast and wheat germ. Nucleic Acids Res., 18, 4677–4682. 31. Kern,D., Roy,H. and Becker,H.D. (2005) The aminoacyl-tRNA synthetases. In Ibba,M, Francklyn,C and Cusack,S (eds), Landes Bioscience, Georgetown, TX, Chapter 20. 32. Feng,L.F., Tumbula-Hansen,D., Min,B., Namgoong,S., Salazar,J., Orellana,O. and Söll,D. (2005). The aminoacyl-tRNA synthetases. In Ibba,M, Francklyn,C and Cusack,S (eds), Landes Bioscience, Georgetown, TX, Chapter 32. 33. Seeman,N.C., Rosenberg,J.M. and Rich,A. (1976) Sequence-specific recognition of double helical nucleic acids by proteins. Proc. Natl. Acad. Sci. USA, 73, 804–808. 34. Wilcox,M. and Nirenberg,M. (1968) Transfert RNA as a cofactor coupling amino acid synthesis with that of protein. Proc. Natl Acad. Sci. USA, 61, 229–236. 35. Lapointe,J., Duplain,L. and Proulx,M. (1986) A single glutamyl-tRNA synthetase aminoacylates tRNA(Glu) and tRNA(Gln) in Bacillus subtilis and efficiently misacylates Escherichia coli tRNA(Gln1) in vitro. J. Bacteriol., 165, 88–93. 36. Stanzel,M., Schön,A. and Sprinzl,M. (1994) Discrimination against misacylated tRNA by chloroplast elongation factor Tu. Eur. J. Biochem., 219, 435–439. 37. Masquida,B. and Westhof,E. (2000) On the wobble GoU and related pairs. RNA, 6, 9–15. 38. Varani,G. and McClain,W.H. (2000) The GU wobble base pair. A fundamental building block of RNA structure crucial to RNA function in diverse biological systems. EMBO Rep., 1, 18–23. 39. Stewart,T.S., Roberts,R.J. and Strominger,J.L. (1971) Novel species of tRNA. Nature, 230, 36–38. 40. Bumsted,R.M., Dahl,J.L., Söll,D. and Strominger,J.L. (1968) Biosynthesis of peptidoglycan of bacterial cell walls. X. Further study of the glycyl transfer ribonucleic acids active in peptidoglycan synthesis in Staphylococcus aureus. J. Biol. Chem., 243, 779–782. 41. Kawakami,M., Tanada,S. and Takemura,S. (1975) Properties of alanyl-oligonucleotide, puromycin, and Staphylococcus epidermidis glycyl-tRNA in interaction with elongation factor Tu:GTP complex. FEBS Lett., 51, 321–324. 42. McClain,W.H. (2006) Surprising contribution to aminoacylation and translation of non-Watson-Crick pairs in tRNA. Proc. Natl Acad. Sci. USA, 103, 4570–4575. 43. RajBhandary,U.L. (1994) Initiator transfer RNAs. J. Bacteriol., 176, 547–552. 44. Dale,T., Sanderson,L.E. and Uhlenbeck,O.C. (2004) The affinity of elongation factor Tu for an aminoacyl-tRNA is modulated by the esterified amino acid. Biochemistry, 43, 6159–6166. 45. Sprinzl,M., Dank,N., Nock,S. and Schön,A. (1991) Compilation of tRNA sequences and sequences of tRNA genes. Nucleic Acids Res., 19(Suppl), 2127–2171. 46. Siatecka,M., Rozek,M., Barciszewski,J. and Mirande,M. (1998) Modular evolution of the Glx-tRNA synthetase family–rooting of the evolutionary tree between the bacteria and archaea/eukarya branches. Eur. J. Biochem., 256, 80–87. 47. Shiba,K., Motegi,H., Yoshida,M. and Noda,T. (1998) Human asparaginyl-tRNA synthetase: molecular cloning and the inference of the evolutionary history of Asx-tRNA synthetase family. Nucleic Acids Res., 26, 5045–5051. 3430 Nucleic Acids Research, 2007, Vol. 35, No. 10 48. Sprinzl,M., Horn,C., Brown,M., Ioudovitch,A. and Steinberg,S. (1998) Compilation of tRNA sequences and sequences of tRNA genes. Nucleic Acids Res., 26, 148–153. 49. Sprinzl,M. and Vassilenko,K.S. (2005) Compilation of tRNA sequences and sequences of tRNA genes. Nucleic Acids Res., 33, D139–D140.